Engagement-Based Segmentation

You can't treat every customer the same. Does every customer deserve a great experience? Absolutely.  Should every customer have the same engagement model? Absolutely not. I said it. I'll die on this hill. Something I've seen in many Series A/B companies, the customer engagement model is the same for a $5K customer as a $500K customer. Same onboarding. Same check-in cadence. Same QBR format. Small customers are over-serviced—too many meetings, too formal for their needs.Teams are on a literal hamster wheel. Large customers are under-served—not enough strategic partnership, you can’t get the execs into a conversation because you’re not having the right conversations or delivering the right value. CS teams are exhausted trying to be everything to everyone. And efficiency is in the toilet. This approach isn't sustainable but many companies default into it while waiting for the “right” time to tackle it. Ready is a decision, not a feeling. Every customer deserves the right engagement model to maximize the value of their investment in your product. But that model has to vary. Different customer sizes, complexity levels, maturity stages, and industries have fundamentally different needs. And economically, it doesn't make sense to deliver the same experience across the board. When you get honest about segmentation, everything changes. In one company I worked with, this is how we approached the first phase of segmentation—we kept it simple: Strategic Accounts (Top 20% of ARR): Named CSM with <30 accounts. Quarterly business reviews with executive sponsors. Custom success plans tied to their business goals. Proactive roadmap discussions. Growth Accounts (Next 30% of ARR): Named CSM with ~60 accounts. Digital engagement supplemented with personal touch. Bi-annual strategic check-ins. Standardized playbooks with customization. Scale Accounts (Remaining 50% of ARR): Pooled support with specialized experts. Digital-first engagement. Automated health monitoring with human escalation when triggered by risk or opportunity. We made the changes and we made no excuses. Customers appreciated the honesty. In the company I mentioned above, customer satisfaction improved across ALL segments. Strategic account retention hit 97%. Scale account retention improved from 86% to 91%. CS costs as a percentage of revenue dropped 35%. CS team engagement scores went up. They were no longer context switching all day every day. Your customer engagement model should be developed and iterated based on what actually works for each customer group. Segmentation isn't about treating customers unfairly—it's about serving them appropriately so each one can achieve maximum value. The model you design today won't be the model you need in 18 months. Customer mix changes. Product evolves. Market shifts. Your engagement approach has to evolve with it. #CustomerSuccess #CustomerExperience #CustomerJourney #RevenueGrowth

POST-4/7👉 Email used to be a megaphone. In 2025, it’s a whisper in a very specific ear. Gone are the days when “blast to all” could pass as a strategy. In fact, that approach in 2025 is actively hurting your deliverability. Email Service Providers (ESPs) like Gmail, Yahoo, and Outlook are no longer just evaluating your IP health—they’re scoring your sender behavior at the recipient level. That means if 40% of your list is cold or disengaged, Gmail sees you as the problem—not just the user. ⚠️ Real Consequence: 1. We audited an ecommerce fashion brand with 220K contacts. Over 92K of them hadn’t clicked a single email in 90+ days. Gmail flagged them for bulk spam behavior, and inboxing fell from 78% to 46% overnight. 2. They were running promos weekly. Nothing was technically broken—but nothing was relevant. That’s what got them crushed. What Micro-Segmentation Solves in 2025: ✅ Reduces spam complaints ✅ Increases engagement velocity ✅ Signals positive intent to inbox providers ✅ Unlocks higher revenue per send with smaller cohorts Micro-Segmentation Tactics That Work Now: 1. Behavior-Based Journeys: Forget static tags. If someone viewed winter boots but didn’t buy, your next 3 emails better talk about warmth, snow, or style—not your general spring lookbook. ✅ Klaviyo + Shopify data lets you trigger flow branches based on: Last viewed product category Cart abandonment by SKU group Pages viewed in session (via UTMs or on-site behavior) Pro Tip: Use dynamic content blocks inside campaigns to adjust hero sections based on browse activity without cloning entire flows. 2. Lifecycle Automation by Spend Velocity This isn’t “new vs returning” logic anymore. In 2025, flows shift based on: Time since last order AOV trends SKU replenishment cycles Example: First-time customer who hasn’t returned in 30 days → “2nd purchase incentive” High-value buyer within 7 days → “VIP early access” Customer inactive 60+ days → Winback + dynamic offer block + channel sync suppression 3. AI-Supported Clustering Tools like RetentionX, Lexer, and even Klaviyo’s predictive analytics are now building multi-dimensional customer clusters using: Purchase frequency Channel source Time to second order Category loyalty It’s loyal mid-value buyers who shop monthly but only when free shipping is offered. ✅ What to do: Export these clusters to your ESP Build messaging that maps exactly to their past actions Suppress low responders from paid channels and warm email instead. Ready to Execute? Create 5 foundational micro-segments: 1. High spenders 2. First-time buyers 3. VIPs (CLV > 2.5x avg) 4. Dormant >90 days 5. Active clickers, no conversion Test 2 cadences per segment: VIPs: 4x/month + early access Dormant: 1x/month reactivation with content—not promos Use Recency, Frequency, and Monetary score buckets to tag customers and let your automations react to movement between them. #EmailMarketing #email

Segmentation is one of those concepts that sounds simple until you actually try to do it properly. Most teams start with broad categories like age, location, or gender, but the real insight comes when you start looking at how users act - how often they visit, how recently they engaged, how much value they bring, and which patterns naturally form across those dimensions. The goal of segmentation isn’t to label users, it’s to understand the structure of their behavior. That’s what data-driven segmentation methods allow us to do. K-Means, for example, helps you find natural patterns hidden in behavioral data. You decide how many groups you want to explore, and the algorithm does the heavy lifting, assigning each user to the cluster that best represents their behavior. It’s simple, efficient, and powerful for large datasets where you want to explore engagement trends without predefining who belongs where. When you need to see relationships instead of just results, hierarchical clustering becomes more useful. It builds a tree-like view showing which users are similar and where meaningful divisions exist. You don’t need to commit to a single number of segments. You can cut the tree at different points to explore how granular your understanding should be. It’s particularly helpful for moderate datasets where interpretability matters as much as precision. Then there’s DBSCAN, a method designed for reality - where user behavior is messy, irregular, and full of noise. Unlike K-Means, DBSCAN doesn’t assume clusters are neat or circular. It groups users by density, identifying natural clusters and automatically separating outliers. This makes it especially valuable for complex behavioral or clickstream data where some users behave in ways that don’t fit any conventional pattern. If you want something more business-focused and immediately actionable, RFM segmentation (Recency, Frequency, Monetary) remains a classic for a reason. By scoring how recently and how often users engage, and how much they contribute, you can pinpoint who’s loyal, who’s at risk, and who’s gone silent. It’s simple but effective for linking behavior to ROI and retention strategies. Finally, once you have meaningful segments, classification models can keep them alive. You can train a model to automatically assign new users to the right segment as data flows in, turning segmentation from a static exercise into a living system that adapts as behavior changes.

My biggest priority at Junction is improving renewal conversations. Not by adding more touchpoints. By making every interaction count. Here are three tactics that actually moved retention: Tactic One: Segment Your Book Most CSMs treat all customers the same. Same cadence. Same agenda. Same deck. That's the fastest way to become background noise. Instead, segment your book by outcome they're driving: → Revenue growth customers → Cost savings customers → Efficiency/workflow customers When you group similar outcomes, you stop context switching between completely different value stories. You get in flow with relevant case studies, metrics that matter, and strategic conversations they actually care about. Tactic Two: Mine for Intelligence Not every customer call needs to drive immediate action. Sometimes you're gathering intelligence for the renewal conversation 90 days out. When you hear "gold nuggets" like: → Upcoming board priorities → Budget reallocation plans → New executive KPIs → Competitive pressure points You capture them. Then you use those insights to frame your value story around what their CFO actually cares about. Tactic Three: Outcomes, Not Features Your customer messages used to sound like this: "Checking in on adoption metrics and wanted to schedule our quarterly review..." Now they sound like this: "I noticed your team is focused on reducing time-to-market by 30% this quarter. Most ops leaders we work with are facing the same tension: pressure to move faster while maintaining quality and compliance." What's more likely: Your customer is thinking about the business outcome you impact? Or your customer is thinking about your product features? Message accordingly, and engagement increases. --- The shift isn't more customer touches. It's more intelligent customer touches. Stop optimizing for activity volume. Start optimizing for strategic relevance. How are you teaching your CS team to segment, mine intelligence, and lead with outcomes?

I grew my client’s best email month by 148% from €525K to €1,3M. Here’s how I did it: When I first took on this client, I loved his vibe and product. But it was tricky… • They only sell 1 product • My client hates being salesy • They sell in multiple countries using native languages Step 1: Audit Every client I take on goes through a thorough list audit. I audited his list with my 66-point Klaviyo checklist to: • Identify what’s leaking revenue • Pinpoint what’s working to double down • Develop a strategy plan to increase his sales Once done and discussed, we got to work on: Step 2: Building and optimizing flows He had multiple flows set up but only his welcome flow was good. The rest was non-existent or only had 1-2 basic emails. So we redid his flows. When I was done, we instantly saw an increase in email revenue. The secret? I used founder-led content. The emails feel like a 1:1 convo, and most of my time went into the copy because… Design attracts but copy sells. The next step was… Step 3: Proper segmentation For a brand that sells one product, it’s easy to neglect the ones who’ve bought. (especially since it’s not a replenishable product) But my client is working on new products. Neglecting anyone would be a HUGE mistake - it always is. Here’s how I segmented: • Language • Engaged vs non-engaged • Customer vs non-customer All emails are properly targeted and personalized with the right intention. It makes people feel like we’re talking directly to them. But this wasn’t enough as… Step 4: Ramp up campaign volume They weren’t sending campaigns which meant a lot of revenue was left on the table. I started with re-engagement campaigns to identify the engaged segment. (if you don’t know this your deliverability will go down the drain) Then I split the engaged segment into: • Non-customers • Everyone That’s how I ensure we don’t send emails that aren’t relevant to the reader. The more campaigns you send, the more touch points and familiarity you create. Which leads to closer relationships and more sales. Many eCom founders are scared of being annoying if they send too many campaigns. If you only send: 1/ Sales and discounts 2/ Things that people don’t care about Then yeah you’re being annoying. But if you do email marketing the quiet way, people want to hear from you. It wasn’t all sunshine and rainbows. This job was brutal. But our hard work paid off because we… ↳ Topped his best sales day in Oct ↳ Grew total revenue by 69% to 4.51M ↳ Beat his best email month by 148% to 1.30M ↳ Grew campaign revenue from 0 to 414k (in Oct) ↳ Involved his customers in his new product development process And the best part? I’ve only been working with them for 3 months. We’re just getting started. -- If you’re an eCom owner who wants to scale your revenue: I’m looking to work with 2 eCom brands to help them increase their email revenue in the next 60 days. DM me “email” and I’ll get you the details.

Most brands segment based on opens or clicks in the last 90 days. Seems safe, right? Wrong. With Apple’s privacy updates (hello, iOS 18.3!) and increasing inbox automation, “open” ≠ engaged anymore. Here's what’s happening: • Apple Mail Privacy inflates open rates • Bots click links to test safety, skewing engagement • Subscribers “opening” emails don’t always remember who you are Better segmentation fix: • Segment by meaningful action (e.g., purchase, site visit, form completion) • Combine multiple signals (click + browse + purchase intent) • Move beyond simplistic windows like “90 day opened” Stop calling passive viewers “engaged”. Real engagement means they took deliberate action. That’s where revenue lives.

Your “engaged segment” is probably costing you email revenue. I audit accounts all the time where segmentation starts and ends with: - 30-day openers - 60-day clickers - 90-day engaged The brand sends content to everyone they think is active. The problem is that's not segmentation. That's filtering. For all of the focus on deliverability in email marketing the last couple of years, we've lost sight of the fact that other data points help drive incremental revenue for a business. Thankfully it's not hard at all to do more, especially when a brand is running on Klaviyo. That means a TON of data is at their disposal. Three simple fixes I look for first: *Exclude recent purchasers* Stop promoting a product to someone who bought it last week. Set a 7-, 14-, or 30-day exclusion window based on your product cycle. If the order isn’t even fulfilled yet, they shouldn’t be in a promo send. *Separate product subscribers from non-subscribers* If someone is already subscribed, they don’t need the pitch again. They need education, usage reinforcement, loyalty building. Cross sell to another product. *Use RFM to change the message* A one-time buyer and a 10-time buyer should not get the same email. Your “Needs attention” segment is often where the incremental lift lives. These people bought before. They just need a reason to come back. Here’s the part most teams miss: The incremental lift doesn't come from creating 10 new campaigns to different groups of users. That's never worth the cost it takes to write, design and deploy that volume (whether you're running things in-house or with an agency). Here's the key: Use dynamic content. Inside a single send: • VIPs see early access • Rewards members see points • Lapsed buyers see a comeback offer • Recent purchasers see something completely different That’s segmentation that drives revenue without multiplying production. Open and click data can inform targeting. It cannot be your growth strategy. If your segmentation doesn’t reflect how people actually buy, you’re not going to make as much money as you should.

LRFM2: A Data-Driven Model for Customer Segmentation 🚀 Not all customers are equal. Some are loyal advocates, some are new opportunities, and some are slowly slipping away. In the world of precision marketing, identifying these groups mathematically is not just an advantage — it’s a necessity. Let me introduce LRFM2 — an advanced upgrade of the classic RFM model, with 2 extra dimensions to make your segmentation truly actionable. --- 1️⃣ What is LRFM? LRFM stands for: L – Length: Time since the first purchase R – Recency: Time since the last purchase F – Frequency: Number of purchases in a defined period M – Monetary: Total spending value --- 2️⃣ What is LRFM2? LRFM2 adds two more dimensions, based on your business needs: X – Engagement (site visits, email opens, app activity) Y – Satisfaction (NPS score, surveys, feedback) So each customer is represented as: [ L, R, F, M, X, Y ] This vector can be used for: Scoring & ranking Clustering (e.g., K-means) Predictive analytics (e.g., CLV) --- 3️⃣ Example Calculation 📅 Assume today is August 8, 2025 Customer First Purchase Last Purchase Orders Spend ($) Engagement (1–5) Satisfaction (1–5) A Jan 2021 Aug 2025 15 3000 5 5 B Jun 2024 Aug 2025 2 400 2 3 C Mar 2023 Mar 2024 5 1200 1 2 D Jul 2022 Jul 2025 10 2500 4 4 --- 4️⃣ Step-by-step metrics L (Length in months) = Today – First Purchase date R (Recency in months) = Today – Last Purchase date F = Orders in the period M = Total spend X = Engagement score (1–5) Y = Satisfaction score (1–5) Example: A: L=56, R=0, F=15, M=3000, X=5, Y=5 B: L=14, R=0, F=2, M=400, X=2, Y=3 C: L=29, R=17, F=5, M=1200, X=1, Y=2 D: L=37, R=1, F=10, M=2500, X=4, Y=4 --- 5️⃣ Normalization & Scoring (Scale: 1–5) We rescale each metric so 1 = worst and 5 = best. For example, in F (frequency), the highest orders get 5 points. A: [5, 5, 5, 5, 5, 5] → Champion 🏆 B: [2, 5, 1, 1, 2, 3] → New Potential 🌱 C: [3, 1, 3, 3, 1, 2] → At-Risk ⚠️ D: [4, 4, 4, 4, 4, 4] → Loyal 💙 --- 6️⃣ Segmentation Strategies Champion (High L, High F, High M) → Loyalty rewards, upsell New Potential → Education, onboarding campaigns At-Risk → Win-back offers, personalized outreach Loyal → Exclusive benefits, early access --- 7️⃣ Final Thought 💡 A customer is not just a transaction — they are a timeline of behaviors, preferences, and engagement patterns. LRFM2 lets you segment based not only on what customers did, but also what they are likely to do next. #LRFM #CustomerSegmentation #DataScience #CRM #MarketingStrategy #PredictiveAnalytics #Retention #CLV #CustomerValue #BusinessModel #LinkedInArticle

I once opened a client’s GA4 account and saw 42 custom segments. Forty-two. Every channel, every micro-behavior, every theoretical funnel stage, broken into separate views. It looked impressive until we realized no one was using them to make a decision. Here’s the danger: over-segmentation creates the illusion of insight. But in reality? It often paralyzes decision-making. ➞ Here’s what I help clients do instead: 1. Build meaningful segments based on shared behaviors: ↳ Combine user traits with action patterns: Engaged Pricing Viewers (users who viewed pricing + triggered a CTA). ↳ Avoid segmenting by single actions - those lead to noise, not trends. 2. Tie each segment to a business decision: ↳ Ask: What would we do differently if this group grew or shrank? ↳ If the answer is unclear, the segment isn’t useful - it’s decoration. 3. Limit active segments to a small, high-value list: ↳ In most GA4 setups, 5–7 audience segments are enough to drive 80% of reporting clarity and campaign targeting. ↳ The rest? Archive or delete. → You don’t need more segments. → You need more strategic ones. → Because when everything is segmented, nothing is prioritized. As a solo consultant, I’ve seen over-segmentation create more confusion than clarity. The best teams I’ve worked with don’t track everything - they track what matters most. How many GA4 segments do you actually use in your decisions? A) 0–5 B) 6–15 C) More than 15 (and I’m scared to delete them)

Email segmentation isn't just a tactic. It's a MINDSET. 💡 I've seen countless marketers blast their entire list with the same message and wonder why their open rates are in the gutter. But here's the thing: Great email marketing isn't about reaching EVERYONE. It's about reaching the RIGHT people with the RIGHT message at the RIGHT time. It's a form of respect, really. You respect your audience by only sending relevant content. You respect your reputation by not forcing messages where they're unwanted. You respect your results by being strategic, not desperate. So what's the solution? A three-layer segmentation approach that transforms your email program: Layer 1: Engagement-Based Segmentation ✅ • Active (opened/clicked in last 30 days) → Regular sending • Warm (31-90 days) → Reduced frequency, value-focused • Cold (90-180 days) → Re-engagement only • Dormant (180+ days) → Suppress or remove This alone tells ISPs your mail is wanted and valued. Layer 2: Risk Tiering 🚦 Ever notice how one bad apple spoils the bunch? Same with email lists. Isolate higher-risk audiences: • New leads or purchased lists → Separate domain • Low engagers → Cautious, infrequent sending • Promotional content → Isolated sending infrastructure Your main domain stays pristine. Your reputation stays intact. Layer 3: Behavior + Demographics 🎯 Now the fun part - personalization based on: • Purchase behavior (what they buy) • Content interests (what they click) • Lifecycle stage (where they are in journey) The real question? Are you still treating your email list as one massive audience? If so, you're leaving engagement on the table and risking your sender reputation. Remember: In email, precision beats volume every time. Segment with intention. Send with purpose. Watch your results transform.